机器人与激光跟踪仪的坐标系转换方法研究

发布时间：2018-06-09 03:07

本文选题：机器人 + 坐标变换　；参考：《中国测试》2017年11期

【摘要】：激光跟踪仪对工业机器人的末端位姿进行绝对定位精度标定时,需要把测量坐标系和机器人坐标系下的坐标值转换到同一坐标系下,坐标值的比较才有意义。因为机器人运动学模型误差的存在,使得机器人坐标系下的点不能遵循着唯一确定的变换关系变换到测量坐标系,所以坐标系的变换精度不高。在建立齐次变换形式的变换模型之后,通过分析坐标系变换过程中的误差来源,提出结合最小二乘法的RANSAC快速转换算法。从测量样点中随机选择一定量的样点拟合变换模型,在获得多个变换模型之后,利用评判模型选择最优模型。结果表明:相比于常规的求解方法,机器人和激光跟踪仪的坐标转换精度提高3倍。RANSAC算法能有效地降低机器人模型误差对求解变换关系的影响,并且实验过程快速、操作简单。
[Abstract]:When the laser tracker calibrates the absolute positioning accuracy of industrial robot, it is necessary to convert the coordinate values of the measuring coordinate system and the robot coordinate system to the same coordinate system, so it is meaningful to compare the coordinate values. Because of the error of robot kinematics model, the point in the robot coordinate system can not be transformed into the measurement coordinate system according to the unique transformation relation, so the transformation precision of the coordinate system is not high. After establishing the transformation model in the form of homogeneous transformation, by analyzing the error sources in the transformation of coordinate system, a fast transformation algorithm of RANSAC combined with least square method is proposed. A certain number of sample fitting transformation models are randomly selected from the measured samples. After obtaining multiple transformation models, the optimal model is selected by using the evaluation model. The results show that the coordinate conversion accuracy of robot and laser tracker can be improved by 3 times compared with the conventional method. RANSAC algorithm can effectively reduce the influence of robot model error on solving the transformation relationship. The experiment process is fast and the operation is simple.
【作者单位】：广东工业大学;
【基金】：广东省前沿与关键技术创新专项资金(2015B010124001) 广东省高等学校优秀青年教师培养计划(YQ2015056) “广东特支计划”科技青年拔尖人才项目(2014TQ01X212) 东莞市产学研项目(2013509109101)
【分类号】：TN249;TP242

【相似文献】